Sea-ice growth or melt is determined by the heat balance between upward ocean heat flux and the conductive heat flux through the overlying ice cover. Melting or freezing of the ice cover feed back to the density profile of surface waters and therefore influence mixing in the upper layer of the ocean. From two deployments of ice mass-balance buoys, in October 2007 in the Bellingshausen Sea and February 2009 in the Amundsen Sea, we determined the ocean heat flux (OHF) for approximately 2 months at each location, spanning the spring–summer and summer–fall transitions, respectively, at the two locations. For both locations the near bottom of the ice cover was isothermal and both sites showed net bottom melting. Average ocean heat flux over the study period was therefore derived directly from the ice-thickness changes measured by an upward-looking acoustic sensor below the ice. These measurements gave an average OHF of 8 W m^–2 at the Bellingshausen Sea site (spring–summer) and 7 W m^–2 in the Amundsen Sea. At the Bellingshausen Sea site, bi-hourly SeaBird CTD measurements were used to determine the ocean temperature elevation about the freezing point. Together with ice speeds determined from GPS positioning, the OHF was parameterized and gave an average value of ~7 W m^–2 in good agreement with that determined from ice melting. The latter method, however, can be used to give instantaneous values of the OHF and showed excursions exceeding 20 W m^–2. These high values were largely determined by the ice speeds rather than the temperature excursions in the mixed layer. The Amundsen Sea site, however, was located on immobile fast ice so the average OHF determined there from ice melting was used with CTD temperature deviations to instead determine the average speed of the currents under the ice. These were determined as about 0.15 m s^–1.